Adaptive NN Optimal Output Regulation for Uncertain Strict‐Feedback Nonlinear System

ABSTRACT This article investigates an adaptive neural network (NN) output‐feedback optimal output regulation problem for an uncertain strict‐feedback nonlinear system with unmeasurable states and unknown disturbances. Since the system states are unmeasurable, an NN state observer is designed to obtain the unknown states. Based on the NN state observer, an adaptive NN output‐feedback optimal tracking controller is proposed. The proposed NN optimal control method consists of an NN feedforward controller and an optimal output‐feedback controller. The NN feedforward controller is designed by output regulation theory and neural networks, which can transform the nonlinear system in strict‐feedback form into a linearizable feedback nonlinear system. The optimal output‐feedback controller is designed for the error nonlinear systems based on adaptive dynamic programming (ADP) and a two‐player zero‐sum game to minimize the cost function and attenuate disturbances. It is shown that the proposed NN optimal control method ensures that the system is asymptotically stable and minimizes the cost function when the disturbances maximize it. Finally, the numerical simulation is given and shows the effectiveness of the proposed adaptive NN output‐feedback optimal tracking control approach.